EST Systems

[J.W. Harris]

>To: diy_efi at coulomb.eng.ohio-state.edu
>From: jwharris at holli.com (J.W. Harris)
>Subject: Re: EST Systems
>Cc: 
>Bcc: 
>X-Attachments: 
>
>>To: Bob Valentine <ravalent at liii.com>
>>From: jwharris at holli.com (J.W. Harris)
>>Subject: Re: EST Systems
>>
>>
>>>>         Dual LCD displays
>>>
>>>    Dual? Where does the second one go?
>>
>>One 16X2 display for each parameter displayed.
>>Line 1: address and hex value.
>>line 2: bargraph.
>>>
>>>    My original idea was to use a GM EST distributor (ones with the 
>>>locked rotor) and dump in a normal 4 wire module.   Set the physical 
>>>timing really retarded, then just delay the impulse from the pickup to 
>>>the module.   Once I get this so the engine will run ok, then I'll add in 
>>>a MAP sensor to compensate for load.
>>>
>>>   Am I way off, or am I at least close?
>>
>>Although some software I've looked at uses a refrence angle which is 
subtracted? out I don't think this is the way to go with a distributor.  Set 
the EST (no advance) distributor up for about 10 - 15 deg. advance like 
normal.  Each EST pulse received will be a refrence for the next cylinder in 
the firing order.  The S/W basically works like this:
>>
>>Each EST interrupt measure the refrence period (time of this EST pulse - 
time of last)
>>
>>Add this to the time of this pulse to predict when the next one will occur 
(this assumes constant engine speed which is approximately correct.  As a 
second cut measure the acceleration which requires comparing this refrence 
period with the last.)
>>
>>We will use an 8 cyl. example.  So each refrence pulse represents 90 deg. 
of crankshaft rotation.  From some algorithim you know how much advance you 
want (another topic) subtract   refperiod *(advance / 90) (fractional time 
part of a refrence period) from the estimated fall time above to determine 
the time to fire the coil (turn it off or make EST go low).  From this 
subtract the dwell time (I haven't got any numbers yet) to determine the 
risetime of EST.  I'm not sure but some modules might limit dwell time 
internally.
>>
>>Of course you need to do the math correctly and might be able to figure 
out how to store the timing map in a way to avoid the division limit minimum 
dwell time, burn time etc.
>>
>>
>>>
>>>   The other reason I like this is that it could easily be fitted to my 
>>>beater/daily driver '83 Olds 307 for testing purposes.   Also since this 
>>>thing could really use some work on the spark map as it is...
>>
>>Retrofits on GMs are really easy just make sure if you are converting a 
points dist. to electronic that you bypass the resistance wire to ignition.
>>
>>>    Since you work for Delco, can you tell me anything about how the EST 
>>>system works on my '83?  I've got the manuals and diagrams, but I'm still 
>>
>>It's difficult to know how every car works but the description above 
should get most efi hackers started on their own.  All the systems work 
bascially the same.  Some Fords (late 80s SEFI) use a distributor pickup 
that produces one ref pulse of a different width so the ECM can tell 
absolute engine position. THe HC11s timer section makes it really easy to 
measure ref pulses and generate EST ones.  I guess thats why we had it built 
that way :).
>>
>>Later,
>>
>
----------------------------------------------------------------------------
J.W. Harris                                          Delco Electronics Corp.
Project Engineer                                     One Corporate Center
Powertrain Electronics                               M.S. CT40C
                                                     Kokomo, IN 46904-9005
jwharris at kocrsv01.delcoelect.com

Of course, all responses are my own and should not be mistaken for those of
Delco Electronics or General Motors.